Avatar

Ten years ago, who would have thought that self-driving cars would be on the brink of being a reality? What’s most important is that having a vision for a self-driving car is one thing but making it a reality is the toughest part of the journey.

During the last decade there was a lot of emphasis on automation in the networking world. Many companies moved away from a manual command line interface (CLI) configuration to automated network provisioning which has resulted in a huge operations expense savings. If automation was the major driving force for this last decade, what will be the big game changer in the next one?

Like automated network provisioning, there are a lot of technologies that are already truly evolving the network. Whether its cloud, hybrid cloud or on-premise infrastructure, the biggest challenge and opportunity in the networking world lies in network analytics and troubleshooting tools.

There are many tools and services that focus on network analytics and assurance. It is no longer a question of “How can I collect information from my devices?”, it’s a question of “What useful information can I extract from my devices with the least CPU and memory consumption?” One of the ways of doing this is by finding a pattern from the collected data and providing a root cause analysis, which saves a lot of time and money. This type of analysis can help a network administrator understand whether the issue is strictly about the network or an end-device or an application. This is where artificial intelligence (AI) and machine learning (ML) can help. Having this level of granular visibility can be vital to an efficient network operation.

Cisco DNA Assurance provides that high-level visibility and more. It helps IT manage the health of the wired and wireless network along with clients and applications. Cisco DNA Assurance also provides a 360-degree contextual view of user, network and applications that allows it to isolate an issue and tell IT where to focus. The number of points in the network that Cisco touches goes all the way from the data center to Enterprise networks, with partnerships with device manufacturers to show end client statistics. And with the ability to capture application performance, Cisco DNA Assurance can provide true network visibility.

Analytics and troubleshooting are important for end-to-end visibility. But one of the most critical pieces in the network—especially for an Enterprise network—is wireless infrastructure. The reason is because with a wired infrastructure it is possible to get the insights easily since it’s a static environment. With wireless infrastructure the main issue is that the environment can be dynamic, especially when it comes to roaming. The one thing a network administrator does not have control over are the end devices in wireless deployments. Having client analytics and correlating this information to network insights and application visibility allows the administrator to look for a pattern. Once found, a remediation action plan based on root cause analysis becomes part of the workflow that leads to a true self-healing network.

So, what should a good network assurance tool include? We can broadly define it in the five main pillars below:

All rights reserved to Sandra Cifo

Visibility

This doesn’t just mean the traditional raw data and telemetry, but the ability to capture useful information. Visibility should include client, network and application visibility.

Analytics

This is long-term visibility and the ability to go back in time to look at the top network issues. The analytics data should be available to network operators in a manner they can consume easily. For example, instead of showing actual device statistics we can assign ratings for how the different network devices are performing.

Real Time Data

Once they have identified an issue, administrators need the ability to capture and display data in real-time or when there is an anomaly detected. This data can be compared to the key performance indicators (KPIs) which helps an admin find the root cause of an issue in real time.

Proactive Analysis

Why wait for an issue to occur and provide remediation? If we already know a pattern, Cisco DNA Assurance help the network administrator mitigate the issue before it occurs. This is where the AI/ML algorithms running in the background help with proactive analysis by recommending suggested actions.

Self-Healing Networks

This is a feature based on the four previous pillars. When proactive remediation is provided, the network admin trusts that the software will automatically take an action without any user intervention. This means less time focusing on the little things, and more times with an eye on the bigger picture.

Cisco’s network assurance provides all the benefits as described above. With years of experience from Cisco technical assistance center (TAC) to root cause customer issues, we have a key advantage of identifying top customer pain points. When we run AI/ML algorithms on top of this rich data set, suggested actions are made which can proactively help in troubleshooting your network. Every time you accept a suggestion from the assurance engine, it paves the way for a true self-healing network tool and makes the tool smarter!

The road to a self-driving network depends on combining these advanced analytics capabilities with the policy automation of a controller-based SDN to achieve an intent-based network that continuously aligns the network to business requirements

Network assurance is a journey that will rely on administrators to provide feedback on if the proactive and predictive insights are actually useful in their environment. This will help the AI/ML engines to evolve over time and come up with remediations that suits an individual customer environment. We are close to a point where we will start trusting our cars for self-driving features. The networking industry is ready to take the next step towards a self-healing network. Is your network infrastructure ready?

 

Learn more about Cisco DNA Network Assurance.